摘要: The phase vocoder is widely used to provide high-fidelity time stretching or contraction of audio signals such as speech monophonic musical passages. Two problems bedevil the reconstructive this technique. First, frequency estimate usually multi-valued and one does not know how choose which possible frequencies given by analysis use in resynthesis. Second, a sinusoidal component incoming sound always excites several channels analysis; reconstructed sine waves can interfere with each other, giving rise reverberant sound. An improved formulation proposed for first difficulty arise; means phase-locking adjacent resynthesis alleviates second one.